During the normal operation of a computer, integrated circuit devices (ICs) generate significant amounts of heat. This heat must be continuously removed, or the integrated circuit device may overheat, resulting in damage to the device and/or a reduction in operating performance. Cooling devices, such as heat sinks, have been used in conjunction with ICs in order to avoid such overheating. Generally, a passive heat sink in combination with a system fan has provided a relatively cost-effective cooling solution. In recent years, however, the power of ICs has increased exponentially, resulting in a significant increase in the amount of heat generated by these devices, thereby making it extremely difficult to extract heat from these devices in order to cool them.
Heat is typically extracted by coupling a heat spreader and a thermal cap to the electronic device as a heat sink. Heat sinks operate by conducting heat from a processor to the heat sink and then radiating it into the ambient air. The better the transfer of heat between the two surfaces (the processor and the heat sink metal) the better the cooling. Some processors come with heat sinks attached to them directly, or interfaced through a thin and soft layer of thermal paste, ensuring a good transfer of heat between the processor and the heat sink. The thermal paste serves not only to transfer heat but to provide some degree of mechanical compliance to compensate for dimensional changes driven by the high operating temperatures of the devices. However, the paste is a weak link in the thermal path. Attempts to thin this layer have resulted in failure of the layer when it is exposed to dimensional changes. There are some known mechanically compliant solutions but these solutions still rely on paste film somewhere in the path.
Printed circuit boards are constructed of various components, some of which have varying coefficients of thermal expansion (CTE). When the components are heated, this can produce degradation such as cracking in some of the components. This is a common problem with solder balls. In addition to the problem of varying CTEs, the different components are usually of varying heights. This adds additional air gaps between current interfaces and circuits, in addition to the air gaps that exist between the components. Heat conduction across air gaps is generally poor, thereby lessening the effectiveness of the heat sink.
In some applications (silicon carrier, multichip modules), it would be desirable to cool several adjacent chips of different thicknesses and perhaps interspersed with capacitors and other components.
Thus, there is a need for a solution that overcomes these shortcomings.